Conformational transitions of proteins are essential for signaling, activation, generation of mechanical energy, regulation and more. They are controlled by subtle changes in the environment and by binding of small molecules and are therefore best modelled at the atomic resolution. However, simulations using atomically detailed descriptions are expensive. Straightforward simulations are restricted to time scale (microseconds) significantly shorter than many important conformational transition of proteins, protein- protein and protein-nucleotide complexes. Our program focused on the development of theory and algorithms for the study of conformational transitions of biological molecules at the correct time scale and at atomic details. To that end we have developed efficient algorithms to compute reaction paths and have applied them to numerous biophysics problems. In the last few years we have extended our set of tools to include Milestoning, a spatial and temporal coarse graining approach to compute thermodynamic and kinetic parameters associated with a reaction path (or with a reaction space). In the present cycle we propose further advancement to the combined method of reaction path calculations and Milestoning which will include expansion of the reaction space under consideration, evaluation of memory effect, and further software enhancements. We will also extend and deepen our investigations of systems we have looked at in the past (myosin II and hemoglobin) and will investigate new systems in tight collaboration with experiments; (i) HIV RT (in collaboration with Ken Johnson) and (ii) IHF (in collaboration with Anjum Ansari). This tight collaboration is expected to shed new light on mechanisms of enzyme specificity and on regulation of transcription.